AI as Tool Maker's Tool

~610 Words | ~2.5min Read

AI isn’t a silver bullet. It’s not an omni-tool that solves every problem. It’s not the miracle everyone’s been waiting for. But what if we’ve been asking the wrong question about what AI actually is?

Think about a 3D printer. The real value isn’t in how it prints, e.g. additive manufacturing. It’s that you can use it to print anything you imagine, so long as you can describe what you’re trying to do. AI works in a similar way. Its chief value isn’t in being the one magic tool. It’s in being a tool for making all kinds of tools.

You’ve might have already experienced this without realizing it. You ask AI to do something. It doesn’t quite work. So you ask it: “How can I make that prompt better?” And it helps you refine the prompt. You used AI as a tool-maker! It created a better tool, the improved prompt, for doing the thing you wanted in the first place. That’s the pattern. Over time, for repeated tasks, you’re gradually making sharper and sharper tools.

But here’s where it gets interesting. AI excels at transformation tasks, that is changing data from one form to another. Summarization, Translation, Reframing, that kind of thing. It’s not great at computation or repeatable execution. But it can generate the script for you! Need something done repeatedly and reliably? AI can help you build the automation. Need to transform information? AI can do it directly or help you create the workflow to do it better next time.

This means our job has shifted. We need to think deeply about our work. Not just what we do, but why we do it. Drucker called this out in the Effective Executive. He said a Knowledge workers need think through what should be done and why. That includes who will use what we’re producing, and their context! We have to define the task well before we can hope to improve it.

What is it you’re trying to do? Why are you trying to do that? Who’s going to use what you’re making? Only then can you identify where AI fits, where it can create tools to make your work better. Then we look for the spots where we need tools. Transformation tasks that could use AI directly. Computation tasks are where we use AI to generate a repeatable script.

Here’s where things get interesting! Like 3D Printing, you don’t have to be an expert in modelling to print a new useful tool. For AI, you don’t have to have the expertise in an expert framework to be able to use it. You just need to be able to describe it… or have a sound explanation from someone who can. AI can create workflows that make expert frameworks executable.

Take something like Wardley Mapping, or decision facilitation, or any framework you’ve read about but struggle to apply consistently. Gather context. Get the framework write-up. Capture your experience and observations. Find some Online discussions of the patterns and anti-patterns. Then work with AI to translate that context into a step-by-step guiding workflow. You’ve just created a tool that can make you and your team better at strategic thinking. That’s AI as tool-maker’s tool in action.

We’re not just automating existing tasks. We’re entering a new tool-making revolution. The only limit is your imagination—and that’s quite literal here. But you need to understand what the tool is capable of first. Then you can use it to the limits of your imagination. The question isn’t “What can AI do for me?” Instead ask: What tools do I need? And how can AI help me make them?